Curie temperature depths in the Alps and the Po Plain (northern Italy): Comparison with heat flow and seismic tomography data

We report on the spectral analysis of the aeromagnetic residuals of the Alps and the Po Plain (northern Italy) to derive the Curie point depth (CPD), assumed to represent the 550 degrees C isotherm depth. We analysed both the aeromagnetic residuals of northern Italy gathered by Agip (now Eni) and the recent EMAG2 compilation. We used the centroid method on 44 and 96 (respectively) 100 x 100 km(2) windows considering both a random and a fractal magnetization distribution, but found that, at least for the Alps, the fractal model yields unrealistically shallow CPDs. Analyses considering a random magnetization model give CPD5 varying between 12 and 39 km (22 to 24 km on average considering the two data sets) in the Po Plain, representing the Adriatic-African foreland area of the Alps, in substantial agreement with recently reported heat flow values of 60-70 mW/m(2). In the Alps, the Eni data set yields shallow CPDs ranging between 6 and 23 km (13 km on average). EMAG2 analysis basically confirms the "hot" Alpine crust, but reduces it to three 50-100 km wide patches elongated along the chain, where CPD5 vary between 10 and 15 km. Such "hot" Alpine domains occur just north of maximum (50-55 km) crustal thickness zones of the Alps and correspond to low seismic wave velocity anomalies recently documented in the 20-22 to 35-38 km depth interval, whereas no relation is apparent with local geology. Assuming an average crustal thermal conductivity of 2.5 W/m degrees C and a steady-state conductive model, CPDs from the hot zones of the Alps translate into heat flow values of 110-120 mW/m(2), and to a basal heat flow from the mantle exceeding 100 mW/m(2) that is significantly greater than that expected in a mountain range. Thus we conclude that the steady-state conductive model does not apply for the Alps and granitic melts occur at similar to 15-40 km depths, consistently with seismic tomography evidence. (C) 2016 Elsevier Ltd. All rights reserved.